Skip to content
LLMs
Moonshot AI logo

MoonshotAI: Kimi K2.5

by Moonshot AI

MoonshotAI: Kimi K2.5 is a large language model from Moonshot AI. It costs $0.450 per million input tokens and $2.25 per million output tokens. Its context window is 262K tokens.

ReasoningTool callingStructured outputPrompt cachingOpen weightsimage input
Input / 1M tokens
$0.450
Output / 1M tokens
$2.25
Cached input / 1M
$0.070

On repeated prefixes

Context window
262K

tokens

Who serves it cheapest

6 hosts serve Kimi K2.5. Same weights, same API - the price difference is pure margin and routing.

Providers serving MoonshotAI: Kimi K2.5, cheapest first
ProviderInput / 1MOutput / 1MContextThroughputUptime 24h
SiliconFlowCheapestint4$0.450$2.25262K-99.9%
AtlasCloudint4$0.490$2.50262K-99.3%
Novita$0.570$2.85262K-93.6%
Amazon Bedrock$0.600$3.00262K-99.4%
Phala$0.600$3.00262K-98.3%
Venice$0.532$3.32256K-100.0%

The spread between SiliconFlow and Venice is 1.4× for identical weights. Quantization and context limits differ, so check both columns before switching.

Benchmarks

Independent scores published alongside the catalogue.

Coding index
46.8

About Kimi K2.5

Kimi K2.5 is Moonshot AI's native multimodal model, delivering state-of-the-art visual coding capability and a self-directed agent swarm paradigm. Built on Kimi K2 with continued pretraining over approximately 15T mixed...

Specifications

MoonshotAI: Kimi K2.5 specifications
Model IDmoonshotai/kimi-k2.5
ProviderMoonshot AI
Context window262K tokens
Max output236K tokens
Input modalitiestext, image
Output modalitiestext
Knowledge cutoff-
Open weightsYes - moonshotai/Kimi-K2.5
ReleasedJanuary 27, 2026

Cheaper alternatives

Models that cost less than Kimi K2.5 while keeping at least half its context window and every input modality it supports.

Frequently asked

How much does MoonshotAI: Kimi K2.5 cost?

$0.450 per million input tokens and $2.25 per million output tokens. Cached input reads cost $0.070 per million tokens.

What is the context window of MoonshotAI: Kimi K2.5?

262K tokens, with up to 236K tokens of output per request.

Which provider serves MoonshotAI: Kimi K2.5 cheapest?

SiliconFlow at $0.450 per million input tokens - 1.4× cheaper than Venice, the most expensive of the 6 hosts serving it.

Confirm against the source: Moonshot AI official pricing.